Interfacial Processes - Diagnostics€¦ · Interfacial Processes – Diagnostics Robert Kostecki...
Transcript of Interfacial Processes - Diagnostics€¦ · Interfacial Processes – Diagnostics Robert Kostecki...
Interfacial Processes – Diagnostics
Robert KosteckiLawrence Berkeley National Laboratory
Berkeley, California 94720June 8, 2010
This presentation does not contain any proprietary, confidential, or otherwise restricted information
Project ID# ES085
Overview
Timeline• PI participated in the BATT Program
since 1999• Task #1 started on Aug. 30, 2009
completed on Dec. 30, 2008• Task #2 initiated on April 1, 2009
70% completed• Task #3 initiated on May 1 2009
70% completed
Budget• FY11 funding TBD• FY10 funding $520K• FY09 funding $520K • FY08 funding $485K• FY07 funding $320K
Barriers Addressed • Inadequate Li-ion battery energy
density, and calendar/cycle lifetimes for PHV and EV applications
• Electrode impedance that limits power density
Partners• ANL, HQ, LBNL, SUNY, UP, and UU
(BATT Task Group "SEI on Alloys") • M. Doeff, LBNL (BATT Task Group
"Phosphates - LiMnPO4")• John Newman (LBNL/UCB) is the
program lead
Objectives
Establish direct correlations between BATT baseline electrodes interfacial phenomena and surface chemistry, morphology, topology, and degradation mechanisms.
Evaluate and improve the capacity and cycle life of intermetallic anodes and high voltage cathodes Determine physico-chemical properties of the SEI i.e., chemical
composition, reactions kinetics, morphology, ionic/electronic conductivity etc.
Investigate electrocatalytic behavior of intermetallic anodes in organic electrolytes
Provide remedies to interface instability e.g., new alloys and/or structures, electrolyte additives, co-deposition of other metals etc.
Characterize degradation modes, improve SEI long-term stability in high-energy Li-ion systems
Evaluate the effect of surface composition and architecture on electrochemical behavior of the electrode
Milestones
1. Evaluate surface and bulk phenomena in LiMnPO4 cathodes (July 2009) Accomplished on time. Novel experimental methodology developed (O2-
plasma etching) to study in situ interfacial pheneomena on composite LiMnPO4 cathodes.
2. Preliminary results of ellipsometric and optical spectroscopy studies of the SEI formation on Sn and Si thin-film anodes (September 2009). Accomplished on time. Preliminary results corroborate previous AFM studies
suggesting that the SEI layer on tin anodes undergoes continuous reformation during cycling.
3. Characterize surface phenomena on LiMnPO4 composite cathodes (July-2010) Work in progress ~90% completed
4. Identify the structural and surface changes of Al, Si and Sn-containinganodes during cycling working collaboratively with the BATT SEI Group.(September-2010). Work in progress ~70% completed
Study of Li+ Transport in Al Membranes
Approach• Improve and use an electrochemical cell of the Devanathan-Stachurski
type to study mass and charge transfer mechanism in intermetallic anodes.
• Develop a time-dependent mass transport model to validate the observed experimental behavior and determine Li transport parameters in aluminum.
• Characterize surface processes in aluminum anodes
Accomplishments• The mechanism of Li+ transport in aluminum was revealed and quantified.
• Higher rate of lithium diffusion in thicker membranes originates from thestructural damage to Al matrix.
• Mixed electrolyte/solid solution transport mechanism and shorter diffusionlength in LixAl phases contribute to this effect.
• α-LixAl solid solution (0<x<0.1).
• Eutectic mixture of α- and β-LixAl solid solutions (0.1<x<0.5).
• β-LixAl solid solution (0.85<x<1).
• Threshold of the cathodic current at0.17 V tends to shift to 0.35 V duringthe reverse scan.
• Dynamic changes in the mechanismand kinetics of Li insertion in LixAlduring cycling.
Cyclic Voltammetry of Al foil, 0.1 mVs-1
electrochemically accessible
What is the mechanism of Li+ transport in Al?
Study of Li+ Transport in Al Membranes
Study of Li+ Transport in Al Membranes
Devanathan-Stachurski Cell for Studies of Li+ Transport in Al
Li+
Li (reference electrode)
A B
Li (counter
electrode)
Li (counter
electrode)
Li (reference electrode)
Al (working
electrode)
EC:DEC 1:1, 1M LiPF6 EC:DEC 1:1, 1M LiPF6
Li+
Li (reference electrode)
A B
Li (counter
electrode)
Li (counter
electrode)
Li (reference electrode)
Al (working
electrode)
EC:DEC 1:1, 1M LiPF6 EC:DEC 1:1, 1M LiPF6
Half-Cell A – Li+ insertion, Al membrane is polarized at U = 10 mV vs. Li/Li+.Half-Cell B - LixAl oxidation, Al membrane is polarized at constant U = 3 V vs. Li/Li+.
M.A.V Devanathan, L. Stachurski, Proc. Roy. Soc., 270, 90 (1962).
membrane
Metallic Electrode in the Devanathan-Stachurski Cell
• Density of charge carriers on the electrode surface (potential) in cell "A" and "B" is different
• The charge carriers are forced in or away from the space charge region at the surface by an electric field created in the cell "A" and "B"
• Space charge region in metallic electrodes is very thin ca. a few Å
Eredox = φ, EF = φM (at equilibrium)
Membrane Permeation Rate
• Permeation is controlled by diffusion in the bulk - the diffusion constant is independent of thickness
• Anomalies in mass transport e.g., phase transitions, structural changes, molecular interactions affect the response signal.
Proc. R. Soc. London A, 270, 90 (1962)
Typical Current Transient in Cell "B"(coverage on the cathodic side is constant)
Current response in cell “A” (0.01 V)
Current response in cell “B” (3 V)
V. E. Guterman et al, Russian J. Electrochem., 37, 57 (2001), V.E. Guterman and L.N. Mironova, Russian J. Electrochem., 36, 517 (2000).
Study of Li+ Transport in Al Membranes
Al membrane thickness
• The initial increase of the cathodic current in cell “A” corresponds to formation of α-LixAl (0 < x < 0.1) solid solution.
• The onset of formation of the β-LiyAl (0.85 < y < 1) phase is marked by the current peak observed after 7000 s.
• The onset (rise time τ0) of the anodic current in cell “B” corresponds to the oxidation of α-LixAl after Li+ reaches side “B” of the membrane.
• The average composition at the steady state conditions reaches ca. Li0.28Al
Side BSide A
Surface Analysis of Al Membranes
• Deep cracks and crevices form during the course of alloying with Li.• The “A” side of the Al membrane exhibits significant morphology changes. It
consists of pillar-like structures with wide gaps and cracks between them. • The “B”-side of the Al membrane displays only moderate surface
degradation.
Why is the diffusion coefficient thickness dependent!?
• Lithium transport can be treated as diffusion controlled, assuming a constant concentration gradient across the aluminum membrane.
• Formation of β-LixAl (0.85<x<1) is associated with substantial volumetric expansion, which induces local stress and leads to Al crystallites expansion, decrepitation and separation.
Side A
Side A
Study of Li+ Transport in Al Membranes
Summary I
Careful micro- and nano-design and advanced manufacturing methods of intermetallic materials is required to improve their rate performance and stability in Li-ion battery applications.
• The measured diffusion coefficients appear to be in relatively good agreement with the reported D values for lithium in Al.
• Observed anomalies in the thickness dependence of the permeation rate of Li in Al suggest a more complicated transport mechanism.
• The amount of mechanical stress is higher in thicker electrodes, which leads to more structural damage.
• The apparent higher rate of lithium diffusion in thicker membranes originates from the mixed electrolyte/solid solution transport mechanism and shorter diffusion length in solid LixAl phases.
Study of Interfacial Phenomena at Sn Anodes
Approach• Apply in situ and ex situ Raman and FTIR spectroscopy, spectroscopic
ellipsometry, AFM, SEM, HRTEM, and electrochemical techniques to detect and characterize surface processes at intermetallic anodes.
• Use model Sn electrodes to detect and monitor early stages of the SEI formation in various electrolytes.
• Determine the nature and kinetics of surface phenomena and their implications for long–term electrochemical performance of the intermetallic anodes in high-energy Li-ion systems.
Accomplishments• Full assessment of interfacial processes on Sn electrode was completed. In situ studies revealed that an effective SEI layer never forms on polycrystalline Sn in
EC-DEC LiPF6 electrolytes. The mechanisms of interfacial processes were determined and characterized.
• Effective strategies to suppress unwanted surface reactions on Sn electrodes were proposed.
Collaboration with BATT Task Group "SEI on Alloys"
Study of Interfacial Phenomena at Sn Anodes
t(min)0 5 10 15 20 25
E(V)
vs
Li+/
Li
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
200 µA cm-2
100665540
t(h)Li alloying
Electrode potential remains >1.3 V at low current densities
• Effective SEI layer never forms at the surface of Sn in EC/DEC LiPF6 electrolyte Continuous reduction reactions at 1.3-1.5V consume the electrolyte.
• Formation of soluble high molecular weight diethyl-2,5-dioxahexane dicarboxylate (DEOHC) is responsible for coulombic inefficiency of Sn anodes. PF5 catalyzed ring-opening polymerization of EC. Electrochemical and chemical decomposition of DEC.
Galvanostatic Polarization of Sn-foil in 1M LiPF6, EC:DEC (1:2)
Cou
nts
0
5
10
15
20
Fresh EC/DEC LiPF6
t(min)0 5 10 15 20
Cou
nts
0
5
10
15
20
Cycled EC/DEC LiPF6
DEC: M=118.13 g mol-1, ε = 2.83 (25°C)
EC: M=88.06 g mol-1, ε = 89.6 (40°C)
GC-MS of the Electrolyte
Electrolyte after electrolysis
Fresh electrolyte
Diethyl-2,5-dioxahexane dicarboxylate
Ex situ AFM Study of Interfacial Phenomena at Sn-foil Anodes
• The surface layer is electronically insulating• Pin holes and cracks in the SEI layer reach the surface of the Sn electrode
Fresh Sn-foil electrode
After scan from 2.7 -> 0.8 V in 1M LiPF6, EC:DEC (1:2), v = 2 mVs-1
(8 x 8 μm images, black color on conductance images represents non-conductive areas )
Morphology Conductance
In situ Spectroscopic Ellipsometry
Interfacial processes can be sensed and monitored in situ at extremely high sensitivity for well defined samples.
Ellipsometry measures the change in polarization state of light reflected from the surface of a sample. The measured ∆ and Ψ values are related to the ratio of Fresnel reflection coefficients Rp and Rs for p and s-polarized light, respectively.
IγIi
In situ Ellipsometry of Sn-foil Anodes
• Significant changes of Δ and Ψ occur at 2.4 and 1.5 V, which correspond to the initial formation of a surface film and further reduction of the electrolyte
• Continuous reformation of SEI upon cycling gradually shifts Δ and Ψ plots shift to higher and lower values, respectively
E(V) vs Li/Li+1.01.52.02.5
j(µA
cm
-2)
-500
-400
-300
-200
-100
0
1st cycle2nd
3rd
4th
1 mVs-1
1M LiPF6, EC/DEC
Forward scans
E(V) vs Li/Li+1.01.52.02.5
Ψ
37.0
37.5
38.0
38.5
39.0
39.5
40.0
E(V) vs Li/Li+1.01.52.02.5
∆
70
72
74
76
78
80
82
84
86
88
Backward scans
E(V) vs Li/Li+1.0 1.5 2.0 2.5
Ψ
37.0
37.5
38.0
38.5
39.0
39.5
40.0
1st (backward scan)2nd3rd4th
E(V) vs Li/Li+1.0 1.5 2.0 2.5
∆
70
72
74
76
78
80
82
84
86
88
In situ Ellipsometry of Sn-foil Anodes
E(V) vs Li/Li+0.51.01.52.02.5
j (µA
cm
-2)
-25
-20
-15
-10
-5
0
5
1st
2nd
3rd
4th
5th 1mV s-1
1M LiPF6, PC
E(V) vs Li/Li+1.01.52.02.5
Ψ
36
38
40
42
44
46
48
1st2nd3rd4th5th
E(V) vs Li/Li+1.0 1.5 2.0 2.5
Ψ
36
38
40
42
44
46
48
1st2nd3rd4th5th
E(V) vs Li/Li+1.01.52.02.5
∆
75
80
85
90
95
E(V) vs Li/Li+1.0 1.5 2.0 2.5
∆
75
80
85
90
95
1st2nd3rd4th5th
Forward scans Backward scans
• Significant changes of Δ and Ψ at 2.4 and 1.5 V in PC electrolyte reflect different SEI layer thickness and composition
• Continuous reformation of SEI upon cycling gradually shifts Δ and Ψ plots shift to higher values
Study of Interfacial Phenomena at Sn Anodes
E(V) vs Li/Li+0.5 1.0 1.5 2.0 2.5 3.0
j( µA
cm
-2)
-120
-100
-80
-60
-40
-20
0
20
LiPF6/EC/DEC electrolyte
1M LiPF6 EC/Glyme
The Effect of Electrolyte Composition
E(V) vs Li/Li+0.5 1.0 1.5 2.0 2.5 3.0
j( µA
cm
-2)
-120
-100
-80
-60
-40
-20
0
20
LiPF6/EC/DEC electrolyte
1M LiPF6 EC/DEC + 0.1M LiF
E(V) vs Li/Li+0.5 1.0 1.5 2.0 2.5 3.0
j( µA
cm
-2)
-120
-100
-80
-60
-40
-20
0
20
LiPF6/EC/DEC electrolyte
1M LiTFSI EC/DEC
DEC EtO- Li+2e-, 2Li+ CO
EC-Li+DEDOHC
Electrochemical initiation*
+ EtO- Li+
Tarascon et al, J. Pow Sources, 178 (2008), 409
• The mechanism of 1M LiPF6 EC/DEC electrolyte reduction involves reduction of DEC to EtO-
followed by chemical reactions with EC and DEC and PF6
-
• Formation of an effective SEI layer on Sn electrodes can be achieved via electrolyte modification solvent(s) (PC, glyme) (LiTFSI) and salt modifications, and additives e.g., LiF, FEC.
LiPF6 (sol) = LiF(s) or (sol) + PF5 (sol)
PF5
+
PF5 Highly reactive
-
Sloop et al, J. Pow. Sources, 119 (2003), 330
Summary II
Preliminary assessment of interfacial processes on Sn electrode was completed
In situ studies revealed that the nature and kinetics of surface reactions are strongly dependent on the electrode and electrolyte. Effective SEI layer never forms at the surface of polycrystalline Sn in EC-
DEC LiPF6 electrolytes
Sn interface instability in organic carbonate electrolytes can be remedied by careful optimization of the electrolyte composition and use of additives that (re)produce a stable SEI layer.
It is critical for the long-term electrochemical performance of intermetallic anodes to suppress unwanted surface reactions. Coordinated electrode and electrolyte design must be carried out to achieve interfacial stability of Sn anodes in Li-ion battery applications.
Interfacial and Bulk Processes in LiMePO4 Cathodes
Approach• Apply in situ and ex situ Raman and FTIR spectroscopy, and
electrochemical techniques to detect and characterize chemical and structural changes in LiMnPO4, and interfacial processes on composite LiMnPO4 cathodes (HPL).Develop an experimental procedure for surface carbon removal through an O2-
plasma etching process to allow microRaman probing of LiMnPO4 composite cathodes.
Design and construct an spectro-electrochemical cell for in situ FTIR-attenuated total reflectance (ATR) microscopy measurements.
Accomplishments• Bulk and surface characterization of a LiMnPO4 electrode was completed.
LiMnPO4 chemical instability upon delithiation was observed and characterized.
Formation of surface films on LiMnPO4 upon charging was observed and characterized.
cm-1
400 600 800 1000 1200 1400 1600 1800R
aman
Inte
nsity
Pristine electrode
After O2 plasma etch
Etched, charged electrode
Etched, discharged electrode
D-band G-band
*
* **
ooooo
+
LiMnPO4 Li4P2O7
10 20 30 40 50 60 70
10 20 30 40 50 60 10 20 30 40 50 60
Charged 181 mAh/g
Discharged 129 mAh/g
0
0
0
0
0
10 µm 10 µm
10 µm10 µm
10 µm 10 µm
Fresh electrode
MicroRaman images of LiMnPO4 composite electrode after O2-plasma etching
Selected Raman spectra of LiMnPO4 composite electrode
• γsymPO4 vibration mode disappears upon charging(!?)
• Li4P2O7 forms upon LiMnPO4 cycling. MnPO4 is chemically unstable vs. the electrolyte?
* LiMnPO4o Li4P2O7
Interfacial and Bulk Processes in LiMePO4 Cathodes
carbon
Li Li
IR Beam
Ge ATR crystalWorking electrode(sample)
Reference and counter electrodes
Electrolyte
Extra tubing for volume adjustment
(‘ATR Theory and Application’ from Pike Technologies application notes)
FTIR-Attenuated Total Reflectance (ATR) Spectro-electrochemical Cell
electrodesample area probed
Interfacial and Bulk Processes in LiMePO4 Cathodes
• Approach mechanism (not shown) to control distance between the Ge ATR crystal and sample, and probing location.
• ATR sample penetration depth ̴ 1 μm.
cm-1
80010001200140016001800
FTIR
Abs
orba
nce
cm-140060080010001200
FTIR
Abs
orba
nce
In situ FTIR-ATR spectraTransmission FTIR spectra
• MnPO4 forms early at the surface of the electrode during charging.• IR bands characteristic for PO4 group tend to vanish upon charging. FTIR data are in
good agreement with Raman results.• Non uniform surface film forms on surface of LiMnPO4 electrode during charging (no
film was detected on pure carbon black electrode at 4.4 V)
0 mAh/g
37 mAh/g
75 mAh/g
112 mAh/g
149 mAh/g
electrolyte
Electrolyte region PO43- region
LiMnPO4 electrode
charged electrode
charged, from in situ cell (149 mAh/g)
Interfacial and Bulk Processes in LiMePO4 Cathodes
Surface film
Summary III
LixMnPO4 is unstable vs. LiPF6: organic carbonate electrolytes. MnPO4 forms early during charging but then tends to undergo
chemical and structural changes to convert to Li4P2O7 after the full cycle of Li extraction and insertion is completed.
Surface film forms at the exposed LixMnPO4 active material. No surface film was detected on carbon black additive.
Instability of delithiated LixMnPO4 likely precludes this material from achieving commercial viability without developing routes for stabilizing the active material.
Future Work• Continue studies of mass and charge transfer mechanisms at the electrode-
electrolyte interface Develop multi-task spectro-electrochemical cell of the Devanathan-Stachurski type to
study in situ and model kinetics of Li intercalation and diffusion through anode and cathode materials (V. Srinivasan)
Carry out quantitative measurements of the mass and charge transfer across electrode/electrolyte interfaces.
• Design and apply in situ and ex situ experimental methodologies to detect and characterize surface processes in Li-ion intermetallic anodes Comprehensive fundamental in situ spectroscopic ellipsometry in conjunction with AFM
and FTIR/Raman surface analysis studies of the SEI layer formation on model monocrystal Sn and Si electrodes will be carried out
Cooperate with the BATT Interfacial Studies Group to investigate the effect of material structure, morphology, topology on formation of the SEI layer
Investigate correlations between physico-chemical properties of the SEI layer and long-term electrochemical performance of Li-ion electrodes
• Diagnostic evaluation of detrimental phenomena in high-voltage (>4.3V) cathodes Apply in situ and ex situ Raman and FTIR spectroscopy to detect and characterize
surface and bulk processes in high voltage cathodes Evaluate the effect of electrode passive additives and impurities on the electrochemical
performance and long-term stability of the composite cathodes.